[elementze]

Agree with your shared thoughts, with exception of the last statement (copied above...);

Most 'upper' (TOP!) frequencies that a DAC chip offers are usually with diminished capability (when we look at their 'white sheet' docs); and given implemented circuits seldom deliver more/equal to a DAC chips 'white sheet' spec; the top most supported frequencies simply give me an indication as to which frequencies might still deliver 'product excellence'.

It is one thing to support a given bitrate or sampling frequency, (and that nets sales from eager consumers wanting to buy 'the next big thing') - but chasing ones tail for esoteric formats that are unlikely to see much use isn't of much worth/'true value' to end consumers..

Far better off easing the burden on your upsampling hardware, and giving the cable a more reasonable bandwidth to push.. the resultant sound by easing off 'one notch' is generally better realised music.

Some circuits DO push past the white sheet values for the DAC chip (iFi do so notably, for example the iFi Diablo clearly does so.. ); but I wouldn't be seeking 1536khz sampling rates from an older BurrBrown DAC chip (a chip built for music rather than 'bragging rights' rates and supporting of unadopted consumer formats etc)..

And if iFi jumped to ESS designs then they would lose me as a customer overnight.
And I am not directly picking on an ESS chipset here with this statement (heaven forbid!), but rather that iFi had a handful of great engineers (or at least several fingers worth of them..) back when they came up with circuits that could push this BurrBrown DAC chip well past its 'theoretical' specs..

Clever implementation is everything in audio, and it constantly amazes me how much an iFi Diablo smokes a Topping D90 for sound output quality.
Maybe some aspect of the Diablo is using a 'class A' output stage (the D90s' biggest weakness to my ears), but it clearly is in their 'multibit' (top six bits) approach that gives detail and definition where it counts, on top of 'top tier specs' beyond what any given consumers system is likely to resolve.

If chasing 1536khz (or 3072khz etc..) is the goal, then 'yes' - some super new DAC chip will be needed.

Just be aware that for approx fifteen years the market for consumer audio has been 'portable' and not home hifi (most surround amps use 'mid tier' DAC chips and are happy with older designs), and so most modern DAC chips are designed to 'sip power' (or give pin outs for I/V stages for configurability to vary their implementation); and chasing future format numbers and rates seems the intent.

Not saying the latest DAC chips are poor, but that as is (un)common knowledge- a DAC is the whole circuit and not just the chip.
iFi have mastered a specific chip, and this fact is of a tangible benefit to us actual end users wanting musicality.
If we consider that two years of R&D was lost by a few manufacturers attempting to design to utilise the AKM449X parts, and then had to throw away that circuit tuning nearly overnight to make use of what the market could supply them, I wouldn't hesitate to predict that the next couple of years will bring a lot of reviews to popular press where industry professionals can RIGHTLY state 'improved implementation over last years/early versions' more so for the ESS chipsets than any other...

As someone who bought (and sold) a lot of PC sound cards in the early nineties, seeing well specced cards 'sound like digital' and vastly more musical cards that didn't 'chase numbers'- I learned early in the game that spec sheets can reveal aspects of the circuit, sure; but certainly not 'musicality' of the design.

Sure the D90 is a great specc'd DAC- but I reiterate- an iFI Diablo smokes it....

I have a very specific use case in mind: Denafrips Pontus2 DAC, I2S input, streamed via HQPlayer up-sampling. The Pontus 2 R2R DAC internally up samples to 32x PCM anyway, so moving this up-sampling outside the box to HQPlayer I can get a theoretical benefit using better (or at least more computationally intensive) up-sampling algorithms. Both the USB and I2S inputs support this sample rate input. Denafrips has a notoriously poor USB implementation (compared to other custom USB implementations, such as Unison USB w/ Schiit). I find USB in on the Denafrips to be noticeably less resolving than AES or I2S. So I've stayed with I2S as the input. But most DDCs only support 8x PCM on I2S. Finding a streamer / DDC that supports 16x and/or 32x PCM over I2S would be very for this particular use case.

I'm certainly not advocating for up-sampling as the greatest audio option. it is fun, and does interesting things to the music. But it can also have noticable downsides.

 

[Greg121986]

Give me Optibox as a digital transport solution please! I would kill for a reliable and noiseless solution that allows me to go from my server way in the basement to my USB DAC upstairs.

 

[Rowan94]

That's quite the effort you put in there! Does that mean the RJ45 port on the OptiBox is 10G capable?

The RJ45 is not 10G capable.

Will you sell longer optical cables or is there an industry standard for these?

We have no plans for that at the moment

Looks nice! It would be great to hear more about the software. In particular, is it based in the same OS as the Zen Stream, and will both have similar upgrade potential?

We'll reveal more information about the software layer shortly, so stay tuned!

 

[Rowan94]

Question...I am uneducated so "does the lack of L/R balanced outlets matter"?

Just to clarify, do you mean the NEO Stream's 4.4mm output?

 

[MFZZ]

I was referring to balanced 3 pin XLR ports. Now I understand....the 4.4 port is not for a headphone, it is for an adapter cable 4.4 to twin XLR. Maybe it would be easier to not mess with an adapter.

 

[Greg121986]

Let's not get too carried away with the "proprietary" implementation of singlemode fiber as a transport. One is not limited to the fiber cable provided by IFI and there is a standard for this. From the pictures, it appears that IFI is using an SC-SC UPC connector with Singlemode Fiber (very standard). This is easy to find for only a few dollars from massive web-based e-tailers as well as industry standard sources like fs.com.

Will you sell longer optical cables or is there an industry standard for these?

We have no plans for that at the moment

 

[Alcophone]

The RJ45 is not 10G capable.

Thank you. So I'm assuming with currently available products, the 10G capability you advertise for the NEO Stream cannot actually be achieved?

Connecting to your network router via cable rather than relying on Wi-Fi? The NEO Stream includes a unique option alongside its RJ45 and M12 ports – optical LAN. This is our own technology. Connect the OptiBox transceiver to the router, and the NEO Stream to the OptiBox via an optical cable. Both the OptiBox and cable are included.

Benefit from the highest quality digital audio connection possible between the router and the streamer.

• Ultra-fast data speeds of up to 10Gbps
• Bit-perfect transmission of ultra-resolution audio
• Minimal degradation across distances of up to 1km

The LAN signal from the router is regenerated, reclocked and rebalanced by the OptiBox; true galvanic isolation is applied, with zero parasitic capacitance and inductance.

(Source)

At least the "This is our own technology" part sounds like the OptiBox is needed to talk to the NEO Stream, i.e. there's no third party device that supports the protocol used between the NEO Stream and the OptiBox, correct?

Let's not get too carried away with the "proprietary" implementation of singlemode fiber as a transport. One is not limited to the fiber cable provided by IFI and there is a standard for this. From the pictures, it appears that IFI is using an SC-SC UPC connector with Singlemode Fiber (very standard). This is easy to find for only a few dollars from massive web-based e-tailers as well as industry standard sources like fs.com.

Thank you! That's great to hear, I was unfamiliar with this standard. I only knew of the smaller, dual fiber standard (LC, apparently) from my own tinkering with 10G Ethernet over OM3 cables. It seems that the 10G standard for consumer devices is RJ45 or SFP+ and the SFP modules with SC support (to use in an SFP+ cage) appear to be limited to 1G capability. Basically I was wondering whether it might be possible to connect the NEO Stream directly to my optical network switch. Physically it seems to be possible, but it still wouldn't achieve the advertised 10G capability and likely not work at all if iFi is using a proprietary protocol. Which is fine, it just seemed like more engineering effort than going with an industry standard.

 

[Rowan94]

I believe with the Zen Stream, HQPlayer was limited to 384k upscaling. Will the NAA interface on this allow up to the full 786k upscaling from HQPlayer and out of the USB into my DAC?

Yes the NEO Stream will support 786kHz.

Do i read this right, there's no headphone output?

That is correct.

Is 5V/1A enough to power OptiBox? Thanks.

Yes 5V/1A.

I was referring to balanced 3 pin XLR ports. Now I understand....the 4.4 port is not for a headphone, it is for an adapter cable 4.4 to twin XLR. Maybe it would be easier to not mess with an adapter.

Take a look at this picture, this should help

 

[bagwell359]

The Zen streamer had teething issues. This box looks to be much more complex. So I will hang back to see what drops.

Many of these features depend on the internal DAC. Going to have to demo that and if it passes, wonder how long before it's passe?

 

[SCBob]

Yes the NEO Stream will support 786kHz.

That is correct.

Yes 5V/1A.

Take a look at this picture, this should help

Will the USB Input (USB Storage) allow USB connection to a MacBook Pro for music play?

 

[Zaurux]

But surprising :

The NEO Stream’s integral DAC stage takes your streaming enjoyment to another level.

• Four-channel True Native hi-res DAC design – supports PCM 768kHz and DSD 512,
• Powerful 16-core XMOS processor – performs full MQA decoding (to 384kHz)

The inscriptions on the motherboard are :
PCM 384 and DSD 4x (256)

But has there been a firmware update since the marketing photo ?

 

[Rowan94]

Thank you. So I'm assuming with currently available products, the 10G capability you advertise for the NEO Stream cannot actually be achieved?

That's the speed of the optical connection between our Optibox and NEO Stream.

(Source)

At least the "This is our own technology" part sounds like the OptiBox is needed to talk to the NEO Stream, i.e. there's no third party device that supports the protocol used between the NEO Stream and the OptiBox, correct?

I'll double check whether the Optibox is the only interface that talks to the NEO Stream's optical input.

Basically I was wondering whether it might be possible to connect the NEO Stream directly to my optical network switch.

We hear you and will find out if this is possible for you

Will the USB Input (USB Storage) allow USB connection to a MacBook Pro for music play?

I don't believe that this could work but I will get this checked for you

 

[Mau10]

But surprising :



The inscriptions on the motherboard are :
PCM 384 and DSD 4x (256)

But has there been a firmware update since the marketing photo ?

I just pull the trigger on an Ifi NEO stream from audio46, due to the excellent results obtained with my ZEN stream for HQplayer and the Hugo TT2. However If NEO were not able to play 768khz and DSD 512, it would be a big disappointment, even when I know that it is supposed to provide a cleaner USB signal.

 

[ra990]

I just pull the trigger on an Ifi NEO stream from audio46, due to the excellent results obtained with my ZEN stream for HQplayer and the Hugo TT2. However If NEO were not able to play 786khz and DSD 512, it would be a big disappointment, even when I know that it is supposed to provide a cleaner USB signal.

Please let me know how this works out for you! In my experience, the USB input on Chord DACs is not up to the task. I have had my little HQPlayer machine run perfectly with the CMA Fifteen's USB and be very glitchy with the Chord USB at the same exact settings and rates. If it doesn't work, it may not be the NEO's fault. I gave up on HQPlayer over USB with Chord. You'll find in the Chord threads that people prefer to use a USB to BNC converter and bypass the USB on the Chord DAC altogether. There's nothing wrong with it at normal hi res rates, but once you go past the 384k rates, it gets glitchy.

 

[Mau10]

Please let me know how this works out for you! In my experience, the USB input on Chord DACs is not up to the task. I have had my little HQPlayer machine run perfectly with the CMA Fifteen's USB and be very glitchy with the Chord USB at the same exact settings and rates. If it doesn't work, it may not be the NEO's fault. I gave up on HQPlayer over USB with Chord. You'll find in the Chord threads that people prefer to use a USB to BNC converter and bypass the USB on the Chord DAC altogether. There's nothing wrong with it at normal hi res rates, but once you go past the 384k rates, it gets glitchy.

Thanks, that is good to know @ra990, I will do some tests, and if it doesn't work, I guess that I will need to place a SRC-DX in between to solve it. Another reason to make my decision is that I am very curious about the Bit perfect native DSD of the Neo Stream, to play with the HQplayer Delta-Sigma modulators. That is a nice feature for a NAA device. Hopefully the Neo's DAC is up to the task regarding sound quality.